PROJECT SUMMARY Over 65 million people in the U.S. have an eating disorder (ED), and almost 20,000 individuals will die per year from ED-related health complications or suicide. Unhealthy (excessive) exercise is an ED symptom present in approximately 50% of people with EDs. When present, excessive exercise is associated with more severe ED psychopathology, slower rates-of-recovery, and faster rates-of-relapse. If left untreated, individuals with EDs engaging in unhealthy exercise are at greater risk for negative physical and psychiatric health outcomes and excess mortality, which, in turn, increase societal costs associated with healthcare utilization. There is a critical need to use objective measures of physical activity in ?real-time? across the weight spectrum in persons with EDs to understand the function of unhealthy physical activity. The scientific premise that guides the proposed research is that the identification of antecedents and consequences associated with unhealthy exercise among persons with EDs will lead to the development of personalized, just-in-time, mHealth interventions that can be sent to individuals with EDs anytime and anywhere. As a first step toward testing our scientific premise, the objectives of this study are to: 1) quantify the frequency of different forms of unhealthy exercise; 2) identify the function of unhealthy exercise relative to temporal changes in affect; and 3) assess individual variation in the function of unhealthy exercise. We will achieve our objectives through a 7-day ecological momentary assessment (EMA) study of non-treatment seeking women with EDs (N=90) recruited from our active ED registry. Participants will: 1) wear a research-grade Actigraph to measure objective physical activity; 2) self- report restricting and loss-of-control eating episodes through PiLR Health, a free mobile phone app designed to collect EMA data; and 3) respond to 6 daily, semi-random and exercise contingent surveys to measure levels of positive and negative affect through the PiLR Health app. Specific aims include: 1) identify the frequency of different unhealthy exercise patterns among women with EDs using real-time objective assessments of physical activity; and 2) examine moment-to-moment between- and within-person antecedents and consequences of unhealthy exercise among women with EDs using real-time assessments of affect and eating. We aim to identify within-person (i.e., personalized) temporal relationships between affect and unhealthy exercise. Knowing an individual?s triggers for impending unhealthy exercise will allow for individualized treatment efforts that do not take a one-size-fits-all approach. By closely monitoring triggers of unhealthy exercise and current physical activity (via smartwatches), automatic and tailored text messages could be sent to persons with EDs to intervene when disordered exercise has occurred or is likely to occur.